基于人工神经网络的箱体类零件的CAPP系统-机械制造及其自动化专业论文.docxVIP

.. ... ........._‘ 摘要  Y 约37(Jfj 本文结合中国第一拖拉机工程机械公司生产实际 F 研制了基于人工神 经网络的变速箱体零件 CAPP 系统一-GBCAPP 。它以成组技术和人工神经 网络技术为基础，采用半创成式工作原理。系统具有零件信息输入 F 工艺 路线自动生成 F 工序图绘制及数据库管理等程序模块。能输出符合企业标 准的机械加工过程卡片飞是一个较为实用的 CAPP 系统。 通过对大量机械零件的结构特点和工艺特征的分析，系统提出了基于 特征的零件信息模型声建立了零件特征分类库 F 开发了适用于各类零件的 信息描述与输入方法一 lt 基于特征的零件信息描述与输入方法飞利用特 征拼装法绘制零件工序简图;利用人工神经网络技术萝自动生成零件的加 工工艺路线。系统软件采用模块化方式，各模块相互独立 F 便于系统的扩 展和升级:系统用户界面良好 F 编写、调试程序和软件维护十分方便。 本文主要以东方红-75 履带式拖拉机变速箱体为例 F 介绍了 GBCAPP 系统#阐述了系统的工作原理、结构和实现方法 F 并给出了系统的运行实 例。 关键词: CAPP 人工神经网络 工序图 .. J 英文摘要 ABSTRACT 飞 衔 ， In terms of the practical requirement of the First Tractor and Construction Machinery Company，this article develops the box part CAPP system based on Artificial Neural Network GBCAPP. It is based on the Group Technology and Artificial Neural Network and adopts integrated CAPP working principle . This system includes the following program modules: inputting p缸ts information，generating process route automatically and producing process drawing. The system a1so can output the process planning sheets in confom世ty with the enterprise standards. This CAPP system is very practical . In this papeιclass libraries 岛r basic feature is founded，and feature based part level modeling is carried on according to the synthetically analysis to the structure character and process property of many mechanical parts. Finally， Based feaωre part description and input method is developed，which is a method suitable for all kinds of p缸t description and input. This article adopts character spelling drawing method to produce the operational sketch. Using Artificial Neural Network on box part，the CAPP system can generate process route automatically. The design of 8ystem so负ware adopts the module 8tructure. The maintenance and expansion of the 80企ware is convenient. ln this thesis，taking gearbox part used on tract Patten of east red-75 tractor as an example，typical part GBCAPP system is introduced in detail. Principle， structure and